You publish a video about a public controversy. The post remains online, yet its reach collapses. Was it rejected by your audience, demoted by an algorithm, restricted under a platform rule, or affected by a government-backed process? Those possibilities may look identical on an analytics screen, but they call for very different responses.
Do not begin with a verdict about censorship. Begin by identifying who acted, what changed, which rule was invoked, and what evidence you can preserve. That method gives citizens, Dharmic organisations, journalists and independent publishers a better chance of challenging a real restriction without mistaking ordinary volatility for deliberate suppression.
Separate speech, distribution and legal control
Every disputed post passes through three layers of decision-making. Confusing them makes a complaint easy to dismiss and an appeal difficult to win.
- Editorial control: You decide what to publish, how to substantiate it and whether a headline accurately represents the underlying material. A weak claim can fail because readers do not trust it, not because anyone censored it.
- Platform control: A social network decides whether the material may remain online and whether it appears in recommendations, searches or followers’ feeds. Content can therefore remain technically available while becoming difficult to discover.
- State control: A government authority may seek restriction or removal under a legal framework. This layer raises the clearest constitutional questions because the State must connect its action to lawful authority and an authorised ground for restricting speech.
The same item can pass one layer and fail another. A lawful political opinion might comply with Bharatiya law yet violate a private platform’s broader community rules. A post might comply with both and still receive little distribution because a recommendation system predicts that users will not engage with it.
Recommendation systems are not neutral noticeboards. They can rank material using signals such as engagement, user behaviour, advertising interests and platform policies. That creates the possibility of a restriction on visibility without an outright deletion. It also means that a fall in impressions, by itself, cannot prove why the fall happened.
The Constitution does not promise any speaker a particular number of views. Visibility still matters, however, because feeds and recommendations have become major routes into public debate. If political criticism, citizen reporting or a community concern remains online but is systematically kept away from readers, the practical effect can resemble exclusion even when the formal right to publish survives.
Diagnose a reach decline before alleging censorship

A reach graph records an outcome. It does not identify the cause. Before making a public allegation or submitting an appeal, create an incident record while the information is still available.
- Save the exact URL, publication time, text, image or video, and any caption or thumbnail used.
- Capture the analytics screen with its date range, not merely the headline number.
- Save every warning, email, in-app notification and account-status message. Record the wording exactly.
- Check whether the item opens through its direct URL when logged out. Then check whether it appears on the account page, in search and in the feeds of actual followers. Each check tests a different form of availability.
- Record any edit made after publication, including a changed caption, removed link or replaced media file.
- Compare performance only with genuinely comparable material. A short video, a long text post and a link to an external website do not provide a clean like-for-like test.
Classify what happened before deciding what to do next:
- Removal: The item no longer opens, and the platform normally provides or should provide a reason. Appeal the identified rule and the specific passage or media at issue.
- Account restriction: Posting, commenting, monetisation or other account functions have been limited. Document the affected functions separately because restoring a post may not restore the account.
- Discoverability restriction: The item remains accessible but is excluded from recommendations, search or another distribution surface. Ask the platform to identify the affected surface, the governing policy and the duration of the limitation.
- Unexplained underperformance: Reach is lower, but no function or distribution surface can be shown to have changed. Keep monitoring and testing; do not present suspicion as proof.
A useful appeal is narrow. Identify the content, quote the notice, name the rule the platform appears to have applied, explain why the material does not meet that rule, and request human review. If the restriction concerns satire, journalism or public-interest reporting, explain that context in concrete terms. A general accusation that the platform is hostile to Hindus or to Bharat may express genuine frustration, but it does not give a reviewer a specific decision to reverse.
Preserve the evidence before editing or deleting anything. If a government order, police communication, court matter, defamation allegation or threat of prosecution is involved, obtain advice from a qualified Bharatiya lawyer promptly. The legal consequences will depend on the exact communication and facts; a platform appeal is not a substitute for legal advice.
Independent publishers should also reduce their dependence on any single feed. Maintain an owned website, a subscriber channel and an organised archive of original media. This is not a licence to evade a lawful order or a platform sanction. It is a continuity measure: a community’s access to its own lawful reporting should not disappear because one account is suspended or one ranking system changes.
Apply the constitutional test to State restrictions

Bharat’s principal digital framework operates through the Information Technology Act, 2000, and the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. The framework has continued to change: amendments in October 2025 modified procedures concerning removal of unlawful information, while amendments in February 2026 addressed synthetically generated information.
Legal authority is not the end of the inquiry. Article 19(1)(a) guarantees citizens freedom of speech and expression, while Article 19(2) permits reasonable restrictions on specified grounds, including public order, defamation, security of the State, and the sovereignty and integrity of India. The working principle is straightforward: speech is protected by default, while a State-imposed restriction needs a legally recognised justification.
When State action affects a post, ask five questions:
- What exact law or rule authorises the action? A general claim that content is objectionable is not the same as a citation to legal authority.
- What specific expression is alleged to be unlawful? The notice should identify the post, statement, image or recording at issue.
- Which permitted ground supports the restriction? Criticism that embarrasses an authority is not automatically a threat to public order or State security.
- Is the response limited to the alleged harm? Removing one unlawful item, disabling an entire account and suppressing all future distribution are different measures with different consequences.
- Is there a meaningful route to challenge the decision? Record the review or appeal mechanism, the authority responsible and any stated deadline.
The Supreme Court supplied an important boundary in Shreya Singhal v. Union of India (2015). It struck down Section 66A of the IT Act because vague and excessively broad language violated freedom of expression. The judgment also stressed the importance of legally recognised procedures when intermediaries are required to remove content. Its practical lesson is not that online speech can never be restricted. It is that undefined notions of annoyance, discomfort or offensiveness cannot become an unlimited licence to silence unpopular opinions.
A second boundary emerged in September 2024. In Kunal Kamra v. Union of India, the Bombay High Court struck down the 2023 amendment that provided for a government Fact Check Unit to identify allegedly false or misleading information concerning government business. The danger was structural: allowing the government to become the decisive judge of claims about itself weakens the independence needed to assess criticism fairly.
Neither decision turns every moderation dispute into unconstitutional censorship. A private platform may enforce standards that go beyond what Bharatiya law prohibits, and an unlawful threat or defamatory statement does not become protected merely because it appears in a political debate. The cases instead give you a disciplined way to examine State power: demand precision, lawful grounds, proportionate action and a real opportunity to contest the decision.
Handle deepfakes as an evidence problem, not just a content problem

A realistic recording no longer proves that the recorded event occurred. A fabricated video of a political leader, a cloned official voice or an invented recording can travel widely before viewers question it. The harm is not limited to one deceived person. Synthetic media can damage reputations, provoke social tension, influence political judgment and erode confidence in authentic evidence.
The February 2026 amendments added platform obligations concerning synthetically generated information, including realistic AI-generated audio-visual material. Identification and safeguards can help users distinguish authentic records from synthetic ones. The difficult part is preserving a boundary between harmful impersonation and legitimate uses such as parody, satire, art and education.
Before your publication or community account shares sensational audio or video, use a repeatable verification routine:
- Find the earliest available upload rather than relying on a repost with a persuasive caption.
- Preserve the URL, file and claim being made. A disappearing upload should not erase the record of how it was presented.
- Look for the longer recording or surrounding context. A genuine clip can still create a false impression when it has been cut selectively.
- Seek confirmation independent of the account that benefits from the claim. Repetition by many accounts is not independent confirmation when they all copy the same upload.
- If your team created or materially altered the media with AI, label that fact prominently and explain whether the work is satire, illustration, reconstruction or another non-documentary form.
- If authenticity remains unresolved, report that uncertainty rather than publishing the media as proof. Do not let the urgency of a political moment lower the evidentiary standard.
A parody label can help prevent confusion, but it does not automatically settle whether the content is lawful. Context, resemblance, distribution and potential harm may all matter. When real legal exposure is possible, have a lawyer assess the finished material before publication.
For a Dharmic publisher, freedom and satya belong in the same editorial discipline. Defending open debate while circulating a convenient fabrication is self-defeating. It gives platforms a reason to distrust later reporting and makes genuine evidence easier for opponents to dismiss.
Demand accountability from platforms and restraint from government

A credible digital order cannot be built by choosing one unaccountable gatekeeper over another. Platforms should not be able to make consequential moderation decisions without explanation, and the government should not acquire an open-ended power to decide which criticism of itself is true.
Use the following standard when evaluating a platform policy, proposed regulation or individual restriction:
- Specific notice: The affected person should be told what content was acted upon, what rule applied and what consequence followed.
- Visibility transparency: A platform should distinguish deletion, account restriction and reduced distribution. Saying only that a post remains online conceals the practical importance of discoverability.
- Meaningful appeal: The user should be able to submit context and obtain review of the actual decision, especially for journalism, political debate and public-interest reporting.
- Scrutiny of automation: Automated moderation systems should be examined for accuracy, consistency and discriminatory outcomes. A machine-made decision can affect public participation as severely as a human one.
- Narrow State action: A government restriction should identify lawful authority, the prohibited material and the harm being addressed. Political discomfort is not a substitute for a ground recognised by Article 19(2).
- A protected space for legitimate synthetic expression: Rules aimed at deceptive deepfakes should not automatically treat disclosed satire, art, education or parody as misinformation.
Key takeaways
- A fall in reach is evidence that distribution changed, not proof of who caused it or why.
- Identify whether you face content removal, an account restriction, a discoverability limit or ordinary underperformance.
- Preserve URLs, notices, analytics and the original media before editing or appealing.
- Use Article 19’s constitutional test when State action is involved; use policy language and the platform’s review process when the decision is private moderation.
- Verify synthetic media before amplification, and clearly label AI-created material when you publish it.
Your most useful next step is to prepare two one-page procedures before the next controversy: an incident log for platform restrictions and a verification checklist for synthetic media. When a post is challenged, follow the evidence from actor to action to rule to remedy. That is how you defend open Dharmic and national discourse without surrendering judgment to an opaque algorithm, a viral fabrication or unchecked State power.
References


Leave a Reply
You must be logged in to post a comment.